Unlock instant, AI-driven research and patent intelligence for your innovation.

Image style migration method, migration device, and computer-readable storage medium

A computer and style technology, applied in the field of image processing, can solve problems such as time-consuming, inapplicable real-time video processing, complex deep network, etc.

Pending Publication Date: 2019-03-26
BEIJING DAJIA INTERNET INFORMATION TECH CO LTD
View PDF4 Cites 12 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] However, for the above two deep network learning methods currently used, if the local analysis and adjustment of the image is encountered during the image stylization process (for example, the building windows are lit in the night stylization of the city landscape), it requires a very complicated Deep network, which makes the process of using deep network to stylize images more time-consuming, not suitable for real-time video processing

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Image style migration method, migration device, and computer-readable storage medium
  • Image style migration method, migration device, and computer-readable storage medium
  • Image style migration method, migration device, and computer-readable storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0045] The embodiment of the present invention provides an image style migration method, such as figure 1 shown, including:

[0046] Step S102, acquiring a content image, wherein the content image is an image that needs to be locally processed in the target area during style transfer;

[0047] Step S104, segmenting the target area from the content image through the target depth network, and performing local processing on the target area to obtain a first processing result;

[0048] Step S106, based on the first processing result, the style-transferred image of the content image is obtained.

[0049] In the embodiment of the present invention, the purpose of local processing of the target area alone is realized through the target deep network, without complicating the network used for style transfer due to local processing, and the time-consuming technology of traditional methods used for image style transfer is alleviated question.

[0050] In an optional implementation man...

Embodiment 2

[0071] An embodiment of the present invention provides an image style transfer device, such as Figure 4 shown, including:

[0072] An acquisition module 100, configured to acquire a content image, wherein the content image is an image that needs to be locally processed in the target area during style transfer;

[0073] The first processing module 200 is configured to segment the target area from the content image through the target depth network, and perform local processing on the target area to obtain a first processing result;

[0074] The determination module 300 is configured to obtain the style-transferred image of the content image based on the first processing result.

[0075] The image style transfer device provided by the embodiment of the present invention realizes the purpose of locally processing the target area alone through the target deep network, without complicating the network used for style transfer due to local processing, and alleviating the traditional...

Embodiment 3

[0085] An embodiment of the present invention provides an image style transfer device, including:

[0086] processor;

[0087] memory for storing processor-executable instructions;

[0088] Wherein, the processor is configured to execute the migration method in Embodiment 1.

[0089] In the embodiment of the present invention, the processor is configured to execute the migration method of Embodiment 1, that is, to obtain the content image, wherein the content image is an image that needs to be locally processed in the target area in the style migration; The region is segmented from the content image, and the target region is locally processed to obtain a first processing result; based on the first processing result, an image after style transfer of the content image is obtained. In this migration method, the purpose of local processing of the target area is realized through the target deep network, without the need to complicate the network used for style transfer due to loc...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The present application relates to an image style migration method, a migration device, and a computer storage medium. The migration method comprises the following steps of: acquiring a content image,wherein the content image is an image which needs to be processed locally in a target region in a style migration; segmentaing The target region from the content image through the target depth network, and processing the target region locally to obtain the first processing result. Based on the first processing result, obtaining an image of the content image after the style transfer. In the migration method, processing the target region locally by the target depth network, and not complicating the network used for the style migration because of the local processing, which alleviates the time-consuming technical problem of the traditional method used for the image style migration.

Description

technical field [0001] The present application belongs to the technical field of image processing, in particular to an image style transfer method and a transfer device. Background technique [0002] Image stylization (image style transfer) is a very interesting image processing method. Under the premise of a given image style, transform other images into images of this style, such as making the image stylized in oil painting, seasonal style, etc. and so on. Image stylization can use traditional learning methods or deep network learning. Deep network learning has more advantages in terms of accuracy. [0003] The current common deep network learning includes deep learning through descriptive neural network and deep learning through generative neural network. Among them, the deep learning is carried out through the descriptive neural network, that is, the stylization of the content image is realized through backpropagation and iteration from random noise. This method minimi...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06T3/00G06T7/11G06T7/194
CPCG06T7/11G06T7/194G06T2207/20084G06T2207/10024G06T3/04
Inventor 张雷张文波郑文
Owner BEIJING DAJIA INTERNET INFORMATION TECH CO LTD